EFFECT OF RED IMPORTED FIRE ANTS ON REPRODUCTION,

HEALTH, AND SURVIVAL OF NORTHERN BOBWHITES

by

JAMES MICHAEL MUELLER, B.S., M.S.

A DISSERTATION

IN

WILDLIFE SCIENCE

Submitted to the Graduate Faculty of Texas Tech University in

Partial Fulfillment of the Requirements for

the Degree of

DOCTOR OF PHILOSOPHY

Approved

August, 1999

1999 by James M. Mueller

ACKNOWLEDGMENTS

The best research is usually a product of good plaiming, adequate funding, hard

work, and both professional and personal support. I was fortunate to be the recipient of

all these factors.

This study was planned by a team of three. Chapter II was the brainchild of Steve

Demarais, who initially was my major professor for a different study. When our project

tumed sour, he guided me to the tutelage and resources of Brad Dabbert. Brad, Steve,

and I clarified our objectives, evaluated different experimental designs, and developed the

methods that led to the successful assessment of the effects of red imported fire ants on

northern bob whites. I am indebted to Steve for his support during the transition between

projects, while a committee member, and continuing after his departure to Mississippi

State University. Likewise, Brad has been the type of major professor of which most

doctoral students only dream. He has always respected my knowledge, tactfully corrected

and encouraged my shortcomings, promptly returned manuscripts, and found the money

to support this research. Chapters III, IV, and V were the result of Brad's guidance.

Steve and Brad have been great examples to me, and I hope to pass on this legacy to

students of my own.

Other faculty also contributed to the success of my research. My committee

members—Harlan Thorvilson, Danny Pence, Scott McMurry, and Rob

Mitchell—^provided helpful advice and expertise. In addition, David Wester sat down

with me at the beginning of my study to review the design and responded to my plea for

help when I could not figure out the proper statistical analyses for certain data.

I received financial support from several entities. First, the Caesar Kleberg

Endowed Fellowship in Wildlife Conservation, administered through the Department of

Range, Wildlife, and Fisheries Management, provided my stipend. The State of Texas,

through the Texas Imported Fire Ant Research and Management Plan and line item

funding for research on fire ants and noxious bmsh and weed control, funded much of the

research. Additional funding was provided by the Department of Range, Wildlife, and

II

Fisheries Management in the form of a startup grant to Brad Dabbert. The J. F. Welder

Heirs Ranch provided me field housing and free-reign over their beautiful ranch.

American Cyanamid provided the fire ant insecticide used for experimental treatments.

Implementing this project in the heat and humidity of South Texas required tough

individuals. I thank my technicians, Mike Whitson and Hunter Haucke, for their

commitment to collection of quality data under all conditions. There is only one thing

worse than having a flat tire on the way back from a long, hot day of walking after birds;

that is having a flat on the way out and the way back and having only one spare tire.

Mike and Hunter overcame situations like this regularly. When I needed to increase my

sample size, Kyle Fiedler dropped everything to come down and help catch birds. During

that brief 10-day period, we had the coldest April weather, the most rattlesnakes, and the

spiciest food that I ever care to experience, but his help was invaluable. I also thank

Andy Forbes, a fellow graduate student who worked with me throughout 1998. Andy

monitored birds for me during March 1998, so I could retum home and be with my

family. If it were not for that, my baby Abigail would not have been conceived!

Finally, I thank my family for their support of my desire to eam a doctorate. My

parents have always encouraged me to pursue my dreams. My sister, Marylee, and her

husband, Eddie, have been especially supportive of my pursuit for a doctorate. But, most

of all, I thank my beautiful wife, Kay, for all she has sacrificed by my leaving a great job

and following my ambition. Kay kept an even-keel when the project on which I started

got canceled and the new project required me to be away for 16 of 20 months. During

this time, she raised an infant, went through a second pregnancy, started her own

business, worked 2 part-time jobs, and drove to San Antonio every 4-6 weeks so I could

visit with her and Zach when the demands of the project would allow it. Her support

allowed me to successfully complete my doctoral program.

Ill

TABLE OF CONTENTS

ACKNOWLEDGMENTS ii

TABLES yi

FIGURES yii

CHAPTER

I. INTRODUCTION 1

Study Area 2

Literature Cited 3

II. NORTHERN BOB WHITE CHICK MORTALITY CAUSED BY RED

IMPORTED FIRE ANTS 4

Abstract 4

Introduction 5

Methods 6

Results 10

Discussion 13

Management Implications 16

Literature Cited 18 III. EFFECT OF RED IMPORTED FIRE ANTS ON

IMMUNOCOMPETENCE OF NORTHERN BOB WHITE CHICKS 21

Abstract 21

Introduction 21

Methods 23

Results 25

Discussion 26

Literature Cited 28

IV. PLASMA CREATINE KINASE LEVELS ASSOCIATED WITH INCREASED PROBABILITY OF MORTALITY IN WILD NORTHERN BOBWHITE 30

Abstract 30

IV

Introduction 31

Methods 32

Results 34

Discussion 37

Literature Cited 39

V. RELATIONSHIP BETWEEN PLASMA TRIGLYCERIDE, BODY

MASS, AND REPRODUCTION OF NORTHERN BOBWHITES 41

Abstract 41

Introduction 41

Methods 43

Results 44

Discussion 45

Literature Cited 51

TABLES

3.1 Masses at 22 days, sheep red blood cell antibody titers, and wing-web indexes for northern bobwhite chicks (N) from nests (n) treated with a fire ant insecticide and controls, Refugio County, Texas, 1998 26

4.1 Tests of linear relationships between the logit of survival to 6 time periods and plasma CK sampled either after transport to a laboratory or immediately in the field 35

5.1 First clutch sizes, dates of nest initiation, late winter body masses (g), and plasma triglyceride levels (mg/dL) of northem bobwhites in 1997 (« = 17) and 1998 (n = 23), Reftigio County, Texas 45

5.2 R^ and significance of late winter body mass (BM) and plasma triglycerides (TRIG) as predictors of first clutch size and date of nest initiation for northem bobwhites in Refugio County, Texas, 1997-1998 47

5.3 Comparison of northem bobwhite late winter body masses (g) and plasma triglyceride levels (mg/dL) in 1997 and 1998 between a single pasture (Willow Lake Trap [WLT], « = 13 and 4, respectively) and all locations (« = 17 and 23, respectively), Refugio County, Texas 48

VI

FIGURES

2.1 Relationship between the proportion of northem bobwhite broods surviving to 3 weeks and number of red imported fire ants in a bait cup after 30 min on the day after hatch (« = 43) 12

4.1 Relationship between survival to 16 weeks of northem bobwhite hens (n = 61) and concentration of plasma CK for birds sampled in the field 36

5.1 Lack of linear relationship between plasma triglycerides and body mass of northem bobwhites in 1997 and 1998, Refugio County, Texas (1997: R^ = 0.01, F, ,5 = 0.2, P = 0.660; 1998: R^ = 0.12, F, 2, = 2.9, P = 0.105) 46

5.2 Lack of linear relationship between plasma triglycerides of northem bobwhites and sampling date in 1998, Reftigio County, Texas (F, 2, = 0.5, P = 0.507) 49

Vll

CHAPTER I

INTRODUCTION

Northem bobwhite (Colinus virginianus) populations have declined throughout their

range during the last 30 years (Sauer et al. 1997). In a study of northem bobwhite

abundance in Texas counties, Allen et al. (1995) found that population declines were

correlated with the year the county was infested by the red imported fire ant (Solenopsis

invicta). Whether the red imported fire ant caused these declines is unknown. However,

Allen et al. (1995) found that areas treated multiple times during 2 years with a fire ant

insecticide had more northem bobwhite in the second year than untreated areas. Thus, it

appears that the red imported fire ant limits northem bobwhite densities in some way.

The red imported fire ant affects northem bobwhite by killing pipping chicks

(Mitchell 1989) and altering activity of young chicks (Pedersen et al. 1996). In addition,

the red imported fire ant may dramatically alter the abundance of invertebrates, an

important component of northem bobwhite diets, when it colonizes an area (Porter and

Savignano 1990). However, there is no evidence northem bobwhite populations are

significantly impacted by these particular effects. The red imported fire ant could affect

northem bobwhite populations through other means, such as by altering growth or

reproduction of northem bobwhite through physiological effects of envenomation.

This study evaluated potential mechanisms by which red imported fire ants impact

northem bobwhite reproduction, survival, and health. The objectives were to determine

the effects of red imported fire ants on hatching success, growth, immunocompetence,

and survival of northem bobwhite chicks.

I conducted 2 other studies simultaneously using the northem bobwhites that were

radiomarked for the study assessing the effect of red imported fire ants. The first of these

studies assessed the relationship between levels of plasma creatine kinase in trapped or

netted northem bobwhite hens and their subsequent survival. High levels of plasma

creatine kinase indicate potential tissue damage from a condition known as capture

myopathy. The second study assessed the relationship between levels of plasma

triglyceride and the date of nest initiation and reproductive effort.

Study Area

All studies were conducted on the J. F. Welder Heirs Ranch in the Texas Coastal

Prairie (Gould 1975; Reftigio County; 28°28'N, 97°11 'W; elevation = 25 m). Grasslands

were interspersed with live oak {Quercus virginiana) motts, honey mesquite (Prosopis

glandulosa), and huisache {Acacia smallii). Climate was subtropical v^th hot summers,

mild winters, and average aimual rainfall of 98 cm (Guckian 1988). Soils were primarily

fine sandy loam or clay with slopes <1% (Guckian 1988). In 1991-92, red imported fire

ants occurred on the study area at a density of 205 mounds/ha and were predominately

polygyne (Allen 1993, Allen et al. 1995).

Literature Cited

Allen, C. R. 1993. Response of wildlife to red imported fire ant population reductions in the South Texas Coastal Prairie. Thesis, Texas Tech University, Lubbock, Texas.

, R. S. Lutz, and S. Demarais. 1995. Red imported fire ant impacts on northem bobwhite populations. Ecological Applications 5:632-638.

Gould, F. W. 1975. The grasses of Texas. Texas A&M University Press, College Station, Texas.

Guckian, W. J. 1988. Soil survey of Refugio County, Texas. U.S. Department of Agriculture, Soil Conservation Service, Washington, D.C.

Mitchell, M. R. 1989. The effects of imported fire ants on nesting bobwhite quail in Texas. Texas Parks & Wildlife Department, Federal Aid in Wildlife Restoration Project W-108-R, Final Report.

Pedersen, E. K., W. E. Grant, and M. T. Longnecker. 1996. Effects of red imported fire ants on newly-hatched northem bobwhite. Journal of Wildlife Management 60:164-169.

Porter, S. D., and D. A. Savignano. 1990. Invasion of polygyne fire ants decimates native ants and dismpts arthropod community. Ecology 71:2095-2106.

Sauer, J. R., J. E. Hines, G. Gough, I. Thomas, and B. G. Peterjohn. 1997. The North American breeding bird survey results and analysis. Version 96.4. Patuxent Wildlife Research Center, Laurel, Maryland.

CHAPTER II

NORTHERN BOBWHITE CHICK MORTALITY CAUSED

BY RED IMPORTED FIRE ANTS

Abstract

Northem bobwhite (Colinus virginianus) populations have declined throughout their

range during the last 30 years. Change in land management is thought to be a primary

cause of this decline, but the invasion of northem bobwhite habitats by the red imported

fire ant (Solenopsis invicta) also may be involved. I compared hatching success and

subsequent survival of wild northem bobwhite chicks on the Coastal Prairie of Texas in

1997 and 1998 between broods that hatched under natural conditions or following fire ant

suppression treatments. In 1997, the fire ant suppression treatment resulted in a 70%

reduction in the number of red imported fire ants on baits placed in treated nests on the

day after hatch. Using 2 fire ant suppression treatments in 1998 reduced the number of

red imported fire ants on baits by >99%. No year or treatment x year interaction effects

were detected for hatching success or survival, and no treatment effect was detected for

hatching success (P > 0.10). However, the proportion of chicks surviving to 21 days was

higher (P = 0.010) for treated nests (« = 18) than control nests (n = 25; proportions of

broods surviving: 53.5 ± 8.6% [x± SE] vs. 24.7 ± 6.6%; chick survival: 60.1 ± 7.6% vs.

22.0 ± 6.2%). The probability of chick survival decreased (P < 0.001) as an index of red

imported fire ant activity in the nest increased. These results indicate northem bobwhite

chicks can suffer high levels of mortality due to red imported fire ants, which could

explain declines in some northem bobwhite populations following infestations by red

imported fire ants. Current methods for controlling red imported fire ants are expensive

and may last <3 months. Thus, strategies for mitigating the effect of red imported fire

ants to northern bobwhite populations in this area should probably focus on reducing

other mortality factors or increasing productivity.

Introduction

Northem bobwhite populations have declined throughout their range during the last

30 years (Sauer et al. 1997). The major cause of these declines is thought to be loss of

habitat due to changes in land use (Brerman 1991). Recently, Allen et al. (1995)

presented associative and experimental results implicating the red imported fire ant as

another factor that may be causing declines in northem bobwhite populations in areas

where their ranges overlap. They found that the begirmings of population declines of

northem bobwhites in Texas counties were correlated with the years the counties became

infested with red imported fire ants; also, multiple treatments with a fire ant insecticide

during 2 years resulted in areas having more northem bobwhites in the second year than

untreated areas (Allen et al. 1995). Thus, it appears that red imported fire ants can limit

northem bobwhite densities in some way.

Researchers have expressed strongly divergent opinions on the role red imported fire

ants may play in northem bobwhite demographics (Brerman 1991, Allen et al. 1993,

Brennan 1993). Some of these differences may stem from confusion about the different

species of fire ants (Solenopsis spp.; Allen et al. 1993). Older studies (e.g., Stoddard

1931, Travis 1938, Johnson 1961, Lehmann 1984) that reported negligible effects of fire

ants on northem bobwhites were conducted before colonization by red imported fire ants

(Allen et al. 1993). However, recent studies that experimentally addressed the effect of

red imported fire ants on northem bobwhites reported adverse effects (Allen et al. 1995,

Giuliano et al. 1996, Pedersen et al. 1996). Red imported fire ants can directly affect

individual northem bobwhites by killing chicks (Mitchell 1989, Giuliano et al. 1996) and

altering their activity (Pedersen et al. 1996). In addition, red imported fire ant

colonization may decrease the abundance of other invertebrates (Porter and Savignano

1990), an important component of northem bobwhite diets (Hurst 1972, Lehmann

1984:170-172). However, previous studies have not provided evidence that these

particular factors affect northem bobwhite populations.

I hypothesized that the time of hatching was a period in the life history of the

northem bobwhite when it was particularly vulnerable to red imported fire ants. I based

my hypothesis on 2 facts. First, other species of fire ants commonly depredate or

scavenge northem bobwhite chicks in the nest (Stoddard 1931, Travis 1938, Johnson

1961, Lehmarm 1984). Second, in a laboratory study with red imported fire ants, survival

of northem bobwhite chicks decreased followdng exposure to ^50 ants for 60 sec or 200

ants for 15 sec (Giuliano et al. 1996). In a wild northem bobwhite nest, it probably takes

^2 hr from the time the first egg in a clutch begins to hatch until the entire brood has

hatched (Stoddard 1931:36-37), and occasionally much longer (Lehmann 1984:89).

During this period, the chicks may be restricted to the nest and vulnerable to stings by red

imported fire ants. I designed this study to measure the effect of red imported fire ants on

hatching success and subsequent survival of northem bobwhite chicks. I tested the null

hypothesis that decreasing the number of red imported fire ants around northem bobwhite

nests on the Texas Coastal Prairie would not affect northem bobwhite hatching success or

subsequent survival.

Methods

Study Design

Experimental units for the study were clutches of eggs in nests I found by tracking

radiomarked birds. I assigned treatments to each clutch completely at random by using a

coin toss as each nest was found. The treatment for each nest was either suppression of

fire ants or no suppression. The model for the experiment was

};., = H + a, + P; + (ap) . + e ,.

The response variable (Yy^) was either the proportion of eggs that hatched successfully or

the proportion of chicks that survived to 3 weeks of age. Two main effects, year and

treatment, and their interaction were assessed. The 2 years of the study were 1997 and

1998.

Field Techniques

I capmred female northem bobwhites from 1 Febmary through 23 April in 1997 and

1998 primarily with walk-in ftmnel traps (Smith et al. 1981) baited with milo. Each

captured hen was radiomarked with a 6.5-g necklace-style radio transmitter without a

body loop (American Wildlife Enterprises, Montacello, Florida).

From late April until mid-July, I located most radiomarked hens at least once every

other day to find nests. I approached each bird on foot until it flushed or moved away, I

observed it, or I had circled it and determined that it might be on a nest. I never

intentionally flushed hens from nests and rarely did so inadvertently. When each nest was

found, I randomly assigned it a treatment. For the fire ant suppression treatment, I used a

handheld broadcast spreader to apply Amdro® (American Cyanamid Company, Wayne,

New Jersey), a fire ant insecticide bait, at a rate of 1.7 kg/ha. I treated a 60- x 60-m area

centered on the nest. The "no suppression" treatment included the same activities as the

fire ant suppression treatment, but without application of the fire ant insecticide bait. Any

nest found <100 m from another nest during the same year was constrained to receive the

same treatment; this constraint affected <15% of nests. Application of Amdro® at rates

used in this experiment typically results in a 90% decrease in red imported fire ant

abundance (Harlan et al. 1981). Initial measurements of effectiveness of the fire ant

suppression treatment (described below) found many red imported fire ants in several

treated nests. To more closely attain my goal of no ants in treated nests, I broadcast

Amdro® directly over the nest for several seconds during the initial application, and I

applied a second treatment of Amdro® to a 20- x 20-m area centered on the nest 1 week

before the expected hatch date. These modifications affected treatment of 3 nests that

hatched in 1997 and all treated nests in 1998.

I measured the effect of red imported fire ant reduction on hatching success and

survival of chicks to 3 weeks of age; in addition, I obtained limited data on survival of

chicks to 1 day of age. Hatching success was estimated on the day of hatch by examining

eggs and eggshells in the nest. Successfully hatched eggs were easy to distinguish by the

even line of pipping that separated the eggshell cap from the remainder of the eggshell.

Unsuccessfully hatched eggs were either unpipped or partially pipped with dead chicks

inside. In addition, any dead chicks found in or around the nest were classified as an

unsuccessfully hatched egg. On the day after hatch, I often approached hens and used the

stereotypical behavior of hens with chicks to determine the number of broods that

suffered 100% mortality within 1 day. This technique could not provide information

beyond presence or absence of chicks; broods in which nearly all chicks died could not be

distinguished from those that had 100% survival. Three weeks after the hatch, I flushed

and counted the hen and chicks to measure chick survival. I repeated these counts vsdth 1

or 2 observers and up to twice per day until I was confident of brood size.

I used a 30-mL plastic cup baited with a slice of hot dog (approx 1 g; Porter and

Tschinkel 1987) to determine efficacy of the fire ant insecticide treatment and assess red

imported fire ant foraging activity in the control nests. On the day of hatch, I removed all

unhatched eggs and eggshells from the nest to eliminate residual yolk in the eggshells to

which the ants might be attracted. On the morning after the eggs hatched, I placed a bait

cup on its side in the nest. After 30 min, I capped the bait cup and placed it in a freezer

(-20 °C) to kill the ants. Bait-cup sampling was conducted within the thermal range of

maximal foraging by red imported fire ants (22-36°C at depth of 2 cm; Porter and

Tschinkel 1987). I counted the number of red imported fire ants in each cup and stored

all ants in 70%) ethanol for later verification of species identification.

I measured the density of active red imported fire ant mounds on the study area

during August 1997. I counted the number of red imported fire ant mounds in 18 20- x

20-m areas centered on northem bobwhite nests that had received the placebo treatment.

To assess presence of red imported fire ants in all potential mounds, I used a shovel to dig

down to visible soil moisture where ants would be if present.

Data Analysis

I compared the proportions of eggs that hatched successftiUy and chicks that

survived to 3 weeks of age between treatments and years using analysis of variance

(ANOVA). The data were analyzed as a 2 x 2 factorial with years and treatments as main

8

effects and yearxtreatment interaction. The year effect was analyzed first as a fixed effect

(Model I ANOVA; Zar 1996:244) to draw inference to treatment effects during the 2

years of this study, and then analyzed as a random effect (Model lU ANOVA or mixed

model; Zar 1996:247) to draw inference to the probability of treatment effects during any

year. Thus, the denominator for the F-test assessing differences in treatment means when

the year effect was fixed was the mean square error of the model, and the denominator for

this test when the year effect was random was the mean of the sum of squares for the

interaction term. I followed the recommendations of Cochran (1943) and determined that

each nest or brood should receive equal weighting. The analysis was repeated using the

ranks of the proportions due to heteroscedasticity and nonnormality (Zar 1996:269-270).

Calculations were made with SPSS for Windows (Nomsis 1993). I report means of the

proportions of eggs that hatched successfully and chicks that survived to 3 weeks of age;

these means were used for experimental comparisons. I also report mean chick survival

to 3 weeks of age for each treatment as calculated via an equation for cluster sampling

(Cochran 1963:65). I estimated the percentage of northem bobwhite mortality at 21 days

attributed to red imported fire ants (m J at hatching by using the equation

m m^ = ( 1 - - ^ ) X 100,

where m^ = the percent mortality at 21 days for the fire ant suppression group, and m^^^ =

the percent mortality at 21 days for the control group. Means are reported ± 1 standard

error.

I used logistic regression to assess the effect of red imported fire ant foraging

activity on chick survival (PROC LOGISTIC; SAS Institute 1989). My model was

(number of chicks from nest /alive at 3 weeks)/(number of chicks from nest /that

hatched) = (number of red imported fire ants in the bait cup in nest I). I tested for a linear

relation between the logit of chick survival and the number of red imported fire ants in

the bait cup in the nest on the day after the hatch by using a full-reduced model likelihood

ratio chi-square test (SAS Institute 1989).

Results

Precipitation differed tremendously during the 2 years of the study. Rainfall from

March through July 1997 was 165% of the long-term mean; during the same period in

1998, rainfall was only 23%) of the long-term mean (National Oceanic and Atmospheric

Administration 1997, 1998).

I radiomarked 77 northem bobwhite hens in 1997 and 71 in 1998. Of 79 nests found

in 1997, 37 were depredated by vertebrates, 35 hatched, 6 were abandoned, and 1 was

trampled by cattle. In 1998,1 found 87 nests: 44 were depredated by vertebrates, 27

hatched, 13 were abandoned, 1 was lost when the hen died, and 2 were still being

incubated when the study ended. Four of the 62 successful nests in 1997-98 were

disturbed by predators before the number of eggs that hatched could be counted, and 1

nest could not be assigned to a treatment due to its proximity to both treated and untreated

nests. These 5 nests were censored. Of the remaining 57 broods, 43 were successfully

counted at 3-weeks of age to measure survival.

Effects of Red Imported Fire Ants

Hatching success did not differ by treatment (F^^2 ~ 0-98, P = 0.327) or year

(F, 53 = 1.89, P = 0.175), and no treatmentxyear interaction was detected (F, 53 = 0.53,

P = 0.472). In the 35 control nests that were successful, 457 of 483 eggs (94.6%)) hatched

successfully. I found 10 of these unsuccessful eggs pipped or hatched, and observed

dozens of red imported fire ants feeding on each dead chick. In the 22 nests treated with

insecticide, 265 of 289 eggs (91.7%)) hatched successfully. I found 4 of these

unsuccessful eggs pipped and 1 cracked, but observed no red imported fire ants on these

chicks.

Chick survival differed greatly between treatments. The proportion of chicks

surviving to 21 days was higher for treated nests (n= 18) than control nests (n = 25;

10

proportions of broods surviving: 53.5 ± 8.6%) vs. 24.7 ± 6.6%), year analyzed as a fixed

effect, F, 39 = 7.35, P = 0.010; chick survival: 60.1 ± 7.6% vs. 22.0 ± 6.2%). Death of all

chicks occurred in 52%) of broods from control nests and 22%) of broods from treated

nests. The percentage of mortality at 21 days attributed to red imported fire ants at

hatching (mj was 48.8%). For this calculation, I did not include the 2%o of chicks from

control nests that were found dead and covered with red imported fire ants because these

chicks may have died of other causes and inclusion of these mortalities had little effect on

the resultant value. No year effect (F, 39 = 0.26, P = 0.612) or treatmentxyear interaction

(F, 39 = 0.18, P = 0.670) were detected. Results using ranked data were similar. When

the year effect was analyzed as a random effect, results were less conclusive (F, , =

39.91, P = 0.100). The power of this test was low because the study was conducted for

only 2 years, which resulted in only 1 degree of freedom for the denominator in the F-test.

I monitored 12 broods on the day after hatch that were later found to have no chicks alive

at 21 days. Six of 10 control broods and both treated broods in this category had lost all

of their chicks within the first day.

The number of red imported fire ants in bait cups in control nests averaged 89 ± 59

more ants in 1997 (x= 358 ± 43 ants in 1997 and 269 ± 39 ants in 1998). The logit of

chick survival decreased as the number of red imported fire ants in a bait cup increased

(X^i = 95.0, P < 0.001; Figure 2.1). The probability of a chick surviving decreased by

0.5%) with each additional red imported fire ant. Mean red imported fire ant mound

density in August 1997 was 290 ± 60 mounds/ha (range = 50-1,000).

In 1998, 1 nest appeared to have been abandoned due to red imported fire ants. The

nest had been constmcted next to a clump of grass that contained an abandoned red

imported fire ant mound. The ants reoccupied the mound after a 3.5-cm rainfall preceded

by 2 months of drought, and they covered the eggs in the nearby northem bobwhite nest

with soil litter.

11

•OX A A

0.0

X Fire Ant Suppression

A Control

0 X

100 200 M v x / ^ i A /y\

300 400 500

NUMBER OF ANTS

600 700 800

Figure 2.1 Relationship between the proportion of northem bobwhite broods surviving to 3 weeks and number of red imported fire ants in a bait cup after 30 min on the day after hatch (« = 43). Broods from nests treated to suppress red imported fire ants are displayed separately from broods from control nests. The logit of chick survival decreased with increasing numbers of red imported fire ants (P < 0.001). The equation depicted by the line is y = e<o55-o.oo54x)/(i + g(o.55-o.oo54x)) Arrows and associated

numbers denote the number of overlapping values at those points.

12

Discussion

Time of Mortality

The benefit of fire ant suppression in my study probably was experienced by

northem bobwhite chicks only while hatching and for a few hours afterward. Chicks

were led away from the nest by 1 or both parents, were always outside the treated area 24

hr following hatching, and were never relocated in a treated area. Thus, chicks from

control and treated nests used the same environment after leaving the nest. Few chicks in

my study were killed in the nest. After 3 weeks, however, survival of broods from

control nests was less than one-half that of broods from treated nests. Much of the

mortality may have occurred soon after the chicks hatched; this possibility is supported by

my data showing that 60%) of the control broods that suffered 100%) mortality by 21 days

had already experienced this mortality by the day after hatch. The following field

observation also supports this possibility and illustrates that deaths may have resulted

from indirect effects of red imported fire ant stings.

At a nest site with a red imported fire ant mound density of 1,000 mounds/ha, I

found 3 dead chicks (2 still in their eggs) in the nest covered by red imported fire ants. I

observed the hen flush 15m from the nest and captured 1 chick that attempted to hide at

my feet. This chick had 1 eye swollen shut from red imported fire ant stings, yellow

marks on its feet that may have been sting sites, and 1 red imported fire ant still stinging

its foot. Pedersen et al. (1996) reported that red imported fire ants often attacked and

stung newly hatched and 1-day-old northem bobwhite chicks on the eyelids, legs, and

toes; stings to the eyelids caused swelling that obscured vision, and stings to the legs and

feet sometimes caused swelling that prevented normal movement. Undoubtedly, the

probability of a chick surviving decreases if its vision and movement are hindered by red

imported fire ant stings.

Chick Survival

Past studies that estimated northem bobwhite chick survival from repeated counts of

chicks in broods associated with radiomarked adults found survival to 2 weeks of 13%) (n

13

= 5 broods; Cantu and Everett 1982) and 38% (n = 22 broods; DeVos and Mueller 1993)

and survival to 3 weeks of 38%) (n = 59 broods; DeMaso et al. 1997). The study area of

DeVos and Mueller (1993) was within the range of the red imported fire ant. When I

compared my results with these studies, survival of chicks that were protected from red

imported fire ants (60%) was higher, and survival of chicks from control nests (22%) was

within the range of previously reported estimates.

When I assumed survival rates of chicks in my 2 treatment groups were identical

after 21 days, additive mortality of northem bobwhite due to stings from red imported fire

ants at the time of hatching was 38.1 ± 7.7%. My estimate of additive mortality was the

difference in chick survival for the 2 treatments. This difference would have been greater

if indirect mortality from red imported fire ant stings did not occur within 21 days.

Giuliano et al. (1996) observed that northem bobwhite chicks that survived the most

severe exposures to red imported fire ants weighed less than controls after 9 days (the

duration of their experiment), and they speculated that depressed growth could reduce

subsequent survival because chick size may be related to survival. However, the

difference in survival rates could have decreased subsequently due to density-dependent

factors. My data could not resolve if the difference in mortality between the 2 treatment

groups changed after 21 days.

Northem bobwhite frequently attempt >1 nest/season (Rosene 1969:73, Roseberry

and Klimstra 1984:83) and may occasionally raise >1 brood in a season (reviewed by

Guthery and Kuvlesky 1998). These attributes allow the possibility that northem

bobwhite could exhibit compensatory reproduction in response to losing broods rapidly

due to red imported fire ant stings at hatching. However, when an entire brood is killed

by red imported fire ants, it occurs after an investment of >37 days by the hen (average

clutch size of 14 eggs [this study], lay rate of 1 egg/day, 23-day incubation period, and

time to build the nest and lay the first egg that Rosene [1969] estimated at 6-12 days).

Burger et al. (1995) found the interval between successfiil brood rearing and renesting

was about 14 days longer than the interval between nest failure and renesting. Thus, if

red imported fire ants kill only some chicks in a brood, even more time will pass before

14

the hen will attempt to renest. In either case, mortality by red imported fire ants at

hatching is more severe than an equivalent proportion of mortality from nest destmction

because of the greater loss of time for reproduction.

I observed no renesting after the loss of an entire brood of chicks, but I only

monitored hens through July. Guthery et al. (1988) found that nearly half the northem

bobwhite hens in my region of Texas were in laying condition in August; thus, hens had

ample time to produce additional clutches. Hens also may have produced clutches that

were incubated by males (Burger et al. 1995, DeMaso et al. 1997). Further investigation

is needed to determine how time constraints resulting from a limited reproductive season

and a large investment of time in chicks that are killed by red imported fire ants affect

potential compensatory reproduction.

The scope of inference from this study is limited to 2 years at 1 site. Here I consider

the reasonableness of extrapolating these results to other times and areas. The statistical

test of treatment effects in the context of all possible years (i.e., years as a random

variable) was not conclusive (P - 0.100). Because precipitation affects northem

bobwhite populations (Rice et al. 1993) and activity of red imported fire ants (Rhoades

and Davis 1967), it was probably the most important variable in the "year" effect in this

study. Fortuitously, precipitation during this study represented periods of flooding one

year and drought the next. Treatment effects were consistent between these conditions.

Thus, I expect effects of red imported fire ants on hatching northem bobwhites to vary

little under the extremes of precipitation that occur on the Texas Coastal Prairie.

However, many other variables could cause differences in the treatment effect, and my

data were not adequate for supporting the effects I observed over the universality of time.

Expanding the inference from this study to other regions is even more problematic.

Because characteristics of red imported fire ants and northem bobwhite may vary among

locations, replication of this study in other regions is needed to understand how well the

fiinctional relation described in Figure 2.1 applies to other regions.

15

Effectiveness of Treatment

The number of red imported fire ants in bait cups exceeded 200 ants at 3 of the first

9 treated nests in 1997 (Figure 2.1); thus, I added a second treatment to attain better ant

suppression. Changing my methodology probably did not confound results. First, chick

survival for these 3 nests did not differ (P = 0.6) from that of other nests treated in 1997.

Second, chick survival across years was not different (P = 0.4) between nests treated

either once or twice. Thus, while the additional treatment was valuable for suppressing

red imported fire ants at some nests, it did not cause a detectable change in chick survival.

I recommend using 2 applications of fire ant insecticide to suppress red imported fire ants

in research applications to ensure distinct differences in numbers of red imported fire ants

between treatments.

Interpreting Evidence at Nests

I observed the outcome of eggs hatching in 3 nests abandoned by the hen (1 due to a

western diamondback rattlesnake [Crotalus atrox] in the nest, 1 due to the initiation of

brooding of the first chick that hatched days before the others, and 1 due to disturbance).

The death of chicks in these nests would have been attributed to red imported fire ants if

the circumstances had not been understood. These nests were tallied as abandoned nests

in my study and not included in any subsequent analysis. Most of the chicks in 2 of these

broods were found dead near their nests and appeared to have been killed by red imported

fire ants. The third brood would probably have suffered the same fate, but that nest had

been treated with fire ant insecticide and the chicks were seen alive around the nest

without a parent for 2 days. These observations confirm that proving red imported fire

ants are responsible for killing northem bobwhite chicks is difficult when researchers

simply observe hatches (Johnson 1961).

Management Implications

Red imported fire ants have probably degraded suitability of habitat for northem

bobwhites throughout the Texas Coastal Prairie due to the mortality they inflict on

16

northem bobwhite chicks. Current methods of controlling red imported fire ants on large

areas using insecticides are expensive ($20.00 ha"' treatment''; Drees et al. 1996) and may

last <3 months (Apperson et al. 1984). Allen (1993) found that insecticidal control of red

imported fire ants to increase northem bobwhite and white-tailed deer (Odocoileus

virginianus) densities for lease hunting was not warranted economically. However,

insecticidal treatments may play a valuable role in management of threatened and

endangered species (e.g., Lockley 1995). One endangered species that occurs near the

study area that might benefit from insecticidal control of red imported fire ants is the

endangered Attwater's prairie chicken (Tympanuchus cupido attwateri).

Because of the high cost of reducing red imported fire ants directly through

insecticide, I think the best current management for northem bobwhite populations is

limited to mitigating red imported fire ant mortality. Thus, reduction of other mortality

factors and increasing productivity via traditional means has become more critical with

the invasion of northem bobwhite habitat by red imported fire ants. The most cost-

effective solution presently is to manage land so that it provides usable space (Guthery

1997) for northem bobwhites in the form of an interspersion of nesting, brood-rearing,

feeding, loafing, and roosting cover demonstrated for decades to benefit northem

bobwhites (Stoddard 1931, Rosene 1969).

17

Literature Cited

Allen, C. R. 1993. Response of wildlife to red imported fire ant population reductions in the South Texas Coastal Prairie. Thesis, Texas Tech University, Lubbock, Texas.

, R. S. Lutz, and S. Demarais. 1993. What about fire ants and northem bobwhites. Wildlife Society Bulletin 21: 349-351.

_ , , and . 1995. Red imported fire ant impacts on northem bobwhite populations. Ecological Applications 5:632-638.

Apperson, C. S., R. B. Leidy, and E, E. Powell. 1984, Effects of Amdro on the red imported fire ant (Hymenoptera: Formicidae) and some nontarget ant species and persistence of Amdro on a pasture habitat in North Carolina. Joumal of Economic Entomology 77:1012-1018.

Brerman, L. A. 1991. How can we reverse the northem bobwhite population decline? Wildlife Society Bulletin 19:544-555.

. 1993. Fire ants and northem bobwhites: a real problem or a red herring? Wildlife Society Bulletin 21:351-355.

Burger, L. W., Jr., M. R. Ryan, T. V. Dailey, and E. W. Kurzejeski. 1995. Reproductive strategies, success, and mating systems of northem bobwhite in Missouri. Joumal of Wildlife Management 59:417-426.

Cantu, R., and D. D. Everett. 1982. Reproductive success and brood survival of bobwhite quail as affected by grazing practices. National Bobwhite Quail Symposium 2:79-83.

Cochran, W. G. 1943. Analysis of variance for percentages based on unequal numbers. Joumal of the American Statistical Association 38:287-301.

. 1963. Sampling techniques. Second edition. John Wiley & Sons, New York, New York.

DeMaso, S. J., A. D. Peoples, S. A. Cox, and E. S. Parry. 1997. Survival of northem bobwhite chicks in western Oklahoma. Joumal of Wildlife Management 61:846-853.

DeVos, T., and B. S. Mueller. 1993. Reproductive ecology of northem bobwhite in north Florida. National Bobwhite Quail Symposium 3:83-90.

Drees, B. M., C. L. Barr, S. B. Vinson, R. E. Gold, M. E. Merchant, and D. Kostroun. 1996. Managing red imported fire ants in urban areas. Texas Agricultural Extension Service Publication B-6043.

Giuliano, W. M., C. R. Allen, R. S. Lutz, and S. Demarais. 1996. Effects of red imported fire ants on northem bobwhite chicks. Joumal of Wildlife Management 60:309-313.

18

Guthery, F. S. 1997. A philosophy of habitat management for northem bobwhites. Joumal of Wildlife Management 61:291-301.

, N. E. Koerth, and D. S. Smith. 1988. Reproduction of northem bobwhites in semiarid environments. Joumal of Wildlife Management 52:144-149.

, and W. P. Kuvlesky, Jr. 1998. The effect of multiple-brooding on age ratios of quail. Joumal of Wildlife Management 62:540-549.

Harlan, D. P., W. A. Banks, H. L. Collins, and C. E. Stringer. 1981. Large area tests of AC-217,300 bait for control of imported fire ants in Alabama, Louisiana, and Texas. Southwestern Entomologist 6:150-157.

Hurst, G. A. 1972. Insects and bobwhite quail brood habitat management. National Bobwhite Quail Symposium 1:65-82.

Johnson, A. S. 1961. Antagonistic relationships between ants and wildlife with special reference to imported fire ants and bobwhite quail in the Southeast. Proceedings of the Aimual Conference of the Southeastem Association of Game and Fish Commissioners 15:88-107.

Lehmann, V. W. 1984. Bobwhites in the Rio Grande Plain of Texas. Texas A&M University Press, College Station, Texas.

Lockley, T. C. 1995. Effect of imported fire ant predation on a population of the least tern—an endangered species. Southwestem Entomologist 20:517-519.

Mitchell, M. R. 1989. The effects of imported fire ants on nesting bobwhite quail in Texas. Texas Parks & Wildlife Department, Federal Aid in Wildlife Restoration Project W-108-R, Final Report.

National Oceanic and Atmospheric Administration. 1997. Climatological data annual summary: Texas. National Climatic Data Center, Asheville, North Carolina.

. 1998. Climatological data annual surrunary: Texas. National Climatic Data Center, Asheville, North Carolina.

Nomsis, M. J. 1993. SPSS® for Windows: base system user's guide. Release 6.0. SPSS, Chicago, Illinois.

Pedersen, E. K., W. E. Grant, and M. T. Longnecker. 1996. Effects of red imported fire ants on newly-hatched northem bobwhite. Joumal of Wildlife Management 60:164-169.

Porter, S. D., and D. A. Savignano. 1990. Invasion of polygyne fire ants decimates native ants and dismpts arthropod community. Ecology 71:2095-2106.

, and W. R. Tschinkel. 1987. Foraging in Solenopsis invicta (Hymenoptera: Formicidae): effects of weather and season. Envirormiental Entomology 16:802-808.

19

Rhoades, W. C , and D. R. Davis. 1967. Effects of meteorological factors on the biology and control of the imported fire ant. Joumal of Economic Entomology 60:554-558.

Rice, S. M., F. S. Guthery, G. S. Spears, S. J. DeMaso, and B. H. Koerth. 1993. A precipitation-habitat model for northem bobwhites on semiarid rangeland. Joumal of Wildlife Management 57:92-102.

Roseberry, J. L., and W. D. Klimstra. 1984. Population ecology of the bobwhite. Southern Illinois University Press, Carbondale, Illinois.

Rosene, W. 1969. The bobwhite quail: its life and management. Rutgers University Press, New Bmnswick, New Jersey.

SAS Institute. 1989. SAS/STAT user's guide. Version 6. Fourth edition. Volume 2. SAS Institute, Gary, North Carolina.

Sauer, J. R., J. E. Hines, G. Gough, I. Thomas, and B. G. Peterjohn. 1997. The North American breeding bird survey results and analysis. Version 96.4. Patuxent Wildlife Research Center, Laurel, Maryland.

Smith, H. D., F. A. Stormer, and R. D. Godfrey, Jr. 1981. A collapsible quail trap. U.S. Forest Service Research Note RM-400.

Stoddard, H. L. 1931. The bobwhite quail: its habits, preservation and increase. Charles Scribner's Sons, New York, New York.

Travis, B. V. 1938. The fire ant (Solenopsis spp.) as a pest of quail. Joumal of Economic Entomology 31:649-652.

Zar, J. H. 1996. Biostatistical analysis. Third edition. Prentice Hall, Upper Saddle River, New Jersey.

20

CHAPTER III

EFFECT OF RED IMPORTED FIRE ANTS ON IMMUNOCOMPETENCE

OF NORTHERN BOBWHITE CHICKS

Abstract

Northem bobwhites can suffer high mortality due to stings from red imported fire

ants while hatching. I conducted a study on wild northem bobwhites to determine if

chicks that survived natural exposure to red imported fire ants at hatching suffered

reduced growth or development of immunocompetence. I compared 22-day chick masses

and measures of humoral and cell-mediated immunity between chicks exposed to natural

levels of red imported fire ants at hatching and chicks from nests where ants were

exterminated. No variables differed between treatments. These results suggest that

growth and immunocompetence of northem bobwhite chicks are probably not affected

greatly by red imported fire ant stings at hatching. Sublethal effects such as reduced

growth can be demonstrated under laboratory conditions, but may occur only rarely under

natural conditions due to the rapid recmitment of red imported fire ants to food sources

and subsequent rapid accumulation of stings by chicks.

Introduction

The red imported fire ant (Solenopsis invicta) is a pest species accidentally

introduced into the United States from South America between 1933 and 1941 (Buren

1972, Rhoades 1977). Its range in the United States rapidly expanded and now

encompasses more than 1 million km^ (Williams 1994). The primary diet of these ants is

insects and other small invertebrates (reviewed in Vinson and Greenberg 1986); however,

they frequently attack and kill vertebrates (reviewed in Allen et al. 1994), including

nestiing wood ducks (Aix sponsa) (Ridlehuber 1982), cliff swallows (Hirundo

pyrrhonota) (Sikes and Amold 1986), least tems (Sterna antillarum) (Lockley 1995), and

northem bobwhites (Chapter II).

21

Before the invasion and spread of red imported fire ants, researchers found that

native fire ants (genus Solenopsis), such as S. geminata and S. richteri, killed 4-9%) of

hatching northem bobwhite (Stoddard 1931, Travis 1938, Johnson 1961). However, the

impact of red imported fire ants to northem bobwhite is more severe than that of native

fire ants. In Chapter II, I estimated that 38.1%) of all northem bobwhites that hatched in

1997-98 died as a result of red imported fire ant stings inflicted at the time of hatching.

This estimate of the proportion of northem bobwhites killed may have been negatively

biased if an assumption, that all mortalities caused by ant stings at hatching occurred

within 21 days, was violated. This assumption was required because chick survival was

measured 21 days after each brood hatched.

Could red imported fire ant stings to chicks at hatching cause mortality >21 days

after hatching? Giuliano et al. (1996) found northem bobwhite chicks that survived the

most severe exposures to red imported fire ants in a laboratory study weighed 12% less

than controls after 9 days (the duration of the experiment). However, 77%) of the chicks

at this level of exposure died (Giuliano et al. 1996). Thus, under laboratory conditions,

most chicks died when exposed to a level of red imported fire ant stings that reduced

growth. Wild northem bobwhites, however, face greater difficulty in acquiring optimal

nutrition, and the effect of red imported fire ant stings may be more severe. For example,

chicks in lab studies might be capable of obtaining optimal nutrition after being partially

blinded by red imported fire ant stings because feed is easily found and provided ad

libitum. Thus, levels of red imported fire ant exposure equivalent to the study by

Giuliano et al. (1996) would probably cause greater mortality and reductions in growth in

wild northem bobwhites; in addition, reductions in growth might occur over a broader

range of red imported fire ant exposure.

Giuliano et al. (1996) did not describe the mechanism by which red imported fire ant

stings reduced growth of northem bobwhite chicks. However, Pedersen et al. (1996)

reported that red imported fire ants often stung the eyelids, legs, and toes of northem

bobwhite chicks. Stings to the eyelid usually caused the eye to swell closed, and stings to

the leg or foot sometimes caused swelling that prevented normal movement. If stings

22

result in reduced intake of protein, chick growth, development of the bursa of Fabricius

and spleen, and cell-mediated immunity might be suppressed (Lochmiller et al. 1993).

Thus, chicks that survive stings from red imported fire ants while hatching may be unable

to acquire an adequate ration of high-protein foods, leading to slower growth, and

resulting in greater susceptibility to pathogens for weeks afterward.

To test whether red imported fire ant stings at hatching could potentially influence

wild northem bobwhite survival >21 days after hatching, I compared 22-day body mass,

cell-mediated immunity, and humoral immunity between northem bobwhite chicks that

hatched from nests with natural levels of red imported fire ants and those from nests

where ants were exterminated. These chicks left the experimentally treated areas around

each nest soon after hatching and thereafter occupied the same natural environment with

uimianipulated red imported fire ant populations.

Methods

Field Techniques

Chapter II provided a detailed description of the experimental treatments. Briefly, I

found nests by monitoring radiomarked northem bobwhite hens from late April until mid

July 1998. When a nest was found, it was randomly assigned to either receive a fire ant

suppression treatment or be a control. For the fire ant suppression treatment, a fire ant

insecticide was applied (1.7 kilograms/ha of Amdro®, American Cyanamid Company,

Wayne, New Jersey) using a handheld broadcast spreader. Treatments were applied to a

60- x 60-m area centered on the nest 3 weeks before the expected hatch date followed by

a retreatment 2 weeks later to the 20- x 20-m core area aroimd the nest. Five variables

were compared between chicks from treated and control nests: hatching success, 21-day

survival, 22-day body mass, humoral immunity, and cell-mediated immunity. Methods

and results for comparisons of hatching success and 21-day survival were reported in

Chapter II. The following describes my methods for capturing chicks and measuring 22-

day body mass, humoral immunity, and cell-mediated immunity.

23

Chick Capture

Broods were captured 22 days after hatching by netting at night. Coveys were found

by tracking the signal from the hen's radio transmitter and illuminating the brood with a

handheld 500,000 candlepower spotlight. Broods were captured using a handheld net

(Mighty Net, Animal Management, Inc., Heafford Junction, Wisconsin; netting was

replaced with 1/4 inch mesh). Chicks were weighed (± 0.1 g), banded with aluminum

legbands, and radiomarked with transmitters weighing 0.7 g (American Wildlife

Enterprises, Montacello, Florida). To assess chick immunocompetence, chicks were

recaptured after 5 and 6 days by locating them using telemetry and capturing them by

hand or net.

Humoral Immunity

Humoral immunity was assessed by measuring antibody production in response to

an injection of sheep red blood cells. Chicks were inoculated intramuscularly with

0.5 mL of 5% sheep red blood cells (Colorado Semm Company, Denver, Colorado) in

phosphate buffered saline (Lochmiller et al. 1993) when initially captured and

radiomarked. Antibody production was assessed after 6 days by pricking the ulnar vein

and collecting blood into capillary tubes. Blood was kept shaded at ambient temperature

for up to 4 hr and then centrifuged for 10 min. Serum was pipetted from the capillary

tubes, placed into cryovials, stored at -20 °C for <9 weeks, shipped on dry ice, and then

stored at -84 °C for 6 months until assayed.

Antibody responses to the sheep red blood cell inoculation was measured using a

microhemagglutination assay (Wegmann and Smithies 1966). As described by

Lochmiller et al. (1993), serial 2-fold dilutions of heat-inactivated semm (56°C for 30

min) in phosphate-buffered saline were mixed with an equal volume of \% sheep red

blood cell solution in phosphate-buffered saline and incubated at 40°C for 1 hr. Titers

were expressed as the logj of the reciprocal of the highest dilution of semm showing a

positive hemagglutination.

24

Cell-mediated Immunitv

Cell-mediated immunity was assessed by measuring wing web inflammation caused

by an intradermal injection of the T-cell mitogen phytohemagglutinin (PHA-P; Sigma, St.

Louis, Missouri). The thickness of each wing web (± 0.01 mm) of 27-day-old chicks was

measured using a digital thickness gauge (Digimatic Thickness Gage, model 547-520,

Mitutoyo Corporation, Takatsu-ku, Kawasaki, Kanagawa, Japan). The right wing web

was injected with 0.5 mg of PHA-P in 0.1 mL of phosphate buffered saline to stimulate

inflammation and the left wing web with 0.1 mL of phosphate buffered saline to serve as

a control (Lochmiller et al. 1993). After 24 hr, the chicks were recaptured and wdng-web

thickness was remeasured. Wing-web swelling was calculated as the difference in wing-

web thickness before and 1 day after injection. The cell-mediated immune response,

termed the wing-web index, was the difference in swelling between the right and left

wing webs (Lochmiller et al. 1993).

Statistical Analysis

Brood means for 22-day chick masses, antibody titers, and wing-web indexes were

compared between treatments using t-tests. Brood means rather than means of all chicks

were used because experimental treatments were applied to nests, not eggs or chicks.

Thus, variables for chicks within a brood were probably correlated. The assumptions of

normality and homoscedasticity were assessed where appropriate using the Shapiro and

Wilk (1965) and Levene (1960) tests, respectively. Analyses were conducted using SPSS

for Windows, release 6.0 (Nomsis 1993) and GPOWER, version 2.0f (Erdfelder et al.

1996). Statistical significance was determined with a = 0.05.

Results

I captured and radiomarked 37 chicks from 11 nests. Of these, 9 lost their radio

transmitters, 6 died, 5 could not be found and were presumed to either have a faulty radio

transmitter or have been depredated, and 17 were successftiUy monitored until they were

28 days old.

25

No differences were detected between chicks that hatched from treated and control

nests for 22-day mass, humoral immunity, or cell-mediated immunity (Table 3.1). Chick

masses ranged from 18.7 to 54.8 g. The heaviest chick failed to express a primary

antibody response to sheep red blood cell immunization and had the third lowest wing-

web index. Excluding this chick, hemagglutinating antibody titers (log2) ranged from 2 to

16 and wing-web indexes ranged from 0.21 to 1.24 mm. Mass was not linearly related to

individual chick antibody titers (F, 9 = 0.13, F = 0.726) or wing-web indexes (F, ,4 =

0.08, F = 0.788).

Table 3.1. Masses at 22 days, sheep red blood cell antibody titers, and wing-web indexes for northem bobwhite chicks (N) from nests (n) treated with a fire ant insecticide and controls, Refugio County, Texas, 1998.

Variable

A

Treated

n X SE N

Control

n X SE / P Power

Mass (g)

Antibody titer (logj)

Wing-web index (mm)

22 6 38.1 2.6

6 4 8.5 2.9

10 5 0.74 0.15

15 5 35.4 3.1 0.69 0.51 0.09

5 3 13.7 2.3 1.32 0.25 0.18

6 3 0.91 0.13 0.77 0.47 0.10

Discussion

Body mass at 22 days and immunocompetence of northem bobwhite chicks were not

affected by exposure to natural levels of red imported fire ants at hatching. Few

comparable data were available to determine the reasonableness of the observed values,

so I calculated an optimal 22-day body mass using an equation derived by Blem and Zara

(1980) for northem bobwhites fed ad libitum on a 30%) protein diet. Blem and Zara

(1980) stated that their asymptotic weight of 205 g was undoubtedly greater than reported

previously for northem bobwhites due to the genetic strain used. For a more realistic

asymptotic weight, I used the average mass of 175 northem bobwhite captured on my

study area in Febmary 1997 and 1998 (x = 169 g). This mean should be a reasonable

estimate of asymptotic weight because body mass of northem bobwhites peaks in

26

Febmary when averaged by gender (Roseberry and Klimstra 1971). Using the other

growth parameters of Blem and Zara (1980) and the site-specific estimate of asymptotic

size, the predicted 22-day body mass of northem bobwhite was 37.9 g. Chicks from

treated nests and control nests attained 100.5%) and 93.4%) of this predicted 22-day body

mass, respectively.

The probability of falsely finding no difference between treatments in body mass and

immunocompetence was high due to the small sample size. For example, the power for

detecting the observed difference in body masses between treatments was 0.09 (Table

3.1). The effect size for this test was 0.41, which is generally considered a small- to

medium-sized effect (Cohen 1977). For a more biologically relevant assessment of

power, I calculated the power of detecting a difference between body mass treatment

means of 10 g. Nestler et al. (1942) reported that 21-day body mass of northem

bobwhites fed either a 22%) or 30%) protein diet differed by about 10 g. The power of

detecting a 10-g difference with the sample sizes (n = 5 and 6), standard deviation (6.7),

and alpha (0.05) of this study was 0.60. Thus, this study had a 60% probability of

detecting an effect of the same magnitude as a 22% versus 30%) protein diet. These

results should be used in context with previous studies to focus future research at the

specific mechanisms by which red imported fire ants impact northem bobwhites. While

my sample size was inadequate for detecting small differences between treatments, my

observed treatment means and previous studies (Giuliano et al. 1996) suggest that growth

and immunocompetence of northem bobwhite chicks are probably not affected greatly by

red imported fire ant stings at hatching. However, northem bobwhite chick survival is

greatly reduced by red imported fire ant stings at hatching (Chapter II).

Under laboratory conditions, the effect of red imported fire ant stings on growth of

northem bobwhite chicks was small and only detectable under conditions that resulted in

77% mortality (Giuliano et al. 1996). Only a narrow range of red imported fire ant

exposure appears to reduce growth without causing death. Thus, I speculate that, under

natural conditions, few hatching northem bobwhite chicks are exposed to red imported

fire ants at a level that is sub-lethal but biologically detrimental.

27

Literature Cited

Allen, C. R., S. Demarais, and R. S. Lutz. 1994. Red imported fire ant impact on wildlife: an overview. Texas Joumal of Science 46:51-59.

Blem, C. R., and J. Zara. 1980. The energetics of young bobwhite (Colinus virginianus). Comparative Biochemistry and Physiology 67A:611-615.

Buren, W. F. 1972. Revisionary studies on the taxonomy of the imported fire ants. Joumal of the Georgia Entomological Society 7:1-26.

Cohen, J. 1977. Statistical power analysis for the behavioral sciences. Revised edition. Academic Press, New York, New York.

Erdfelder, E., Faul, F., and A. Buchner. 1996. GPOWER: A general power analysis program. Behavior Research Methods, Instmments, & Computers 28:1-11.

GiuHano, W. M., C. R. Allen, R. S. Lutz, and S. Demarais. 1996. Effects of red imported fire ants on northem bobwhite chicks. Joumal of Wildlife Management 60:309-313.

Johnson, A. S. 1961. Antagonistic relationships between ants and wildlife with special reference to imported fire ants and bobwhite quail in the Southeast. Proceedings of the Aimual Conference of the Southeastem Association of Game and Fish Commissioners 15:88-107.

Levene, H. 1960. Robust tests for equality of variances. Pages 278-292 in I. Olkin, editor. Contributions to probability and statistics. Stanford University Press, Palo Alto, California.

Lochmiller, R. L., M. R. Vestey, and J. C. Boren. 1993. Relationship between protein nutritional status and immunocompetence in northem bobwhite chicks. Auk 110:503-510.

Lockley, T. C. 1995. Effect of imported fire ant predation on a population of the least tern—an endangered species. Southwestem Entomologist 20:517-519.

Nestler, R. B., W. W. Bailey, and H. E. McCIure. 1942. Protein requirements of bobwhite quail chicks for survival, growth, and efficiency of feed utilization. Journal of Wildlife Management 6:185-193.

Nomsis, M. J. 1993. SPSS® for Windows: base system user's guide. Release 6.0. SPSS, Chicago, Illinois.

Pedersen, E. K., W. E. Grant, and M. T. Longnecker. 1996. Effects of red imported fire ants on newly-hatched northem bobwhite. Joumal of Wildlife Management 60:164-169.

Rhoades, R. B. 1977. Medical aspects of the imported fire ant. University Presses of Florida, Gainesville, Florida.

28

Ridlehuber, K. T. 1982. Fire ant predation on wood duck ducklings and pipped eggs. Southwestem Naturalist 27:222.

Roseberry, J. L., and W. D. Klimstra. 1971. Annual weight cycles in male and female bobwhite quail. Auk 88:116-123.

Shapiro, S. S., and M. B. Wilk. 1965. An analysis of variance test for normality (complete samples). Biometrika 52:591-611.

Sikes, P. J., and K. A. Amold. 1986. Red imported fire ant (Solenopsis invicta) predation on cliff swallows (Hirundo pyrrhonota) nestlings in east-central Texas. Southwestem Naturalist 31:105-106.

Stoddard, H. L. 1931. The bobwhite quail: its habits, preservation and increase. Charles Scribner's Sons, New York, New York.

Travis, B. V. 1938. The fire ant (Solenopsis spp.) as a pest of quail. Joumal of Economic Entomology 31:649-652.

Vinson, S. B., and L. Greenberg. 1986. The biology, physiology, and ecology of imported fire ants. Pages 193-226 in S. B. Vinson, editor. Economic impact and control of social insects. Praeger Publishers, Westport, Connecticut.

Wegmann, T. G., and O. Smithies. 1966. A simple hemagglutination system requiring small amoimts of cells and antibodies. Transfusion 6:67-73.

Williams, D. F. 1994. Control of the introduced pest Solenopsis invicta in the United States. Pages 282-292 in D. F. Williams, editor. Exotic ants: biology, impact, and control of introduced species. Westview Press, Boulder, Colorado.

29

CHAPTER IV

PLASMA CREATINE KINASE LEVELS ASSOCIATED WITH INCREASED

PROBABILITY OF MORTALITY IN WILD NORTHERN BOBWHITE

Abstract

I assessed the effect of capture myopathy on survival of northem bobwhites (Colinus

virginianus) during January through August of 1997 (n = 48) and 1998 (n = 56) in Texas.

I captured bobwhites by trapping and netting and measured plasma creatine kinase (CK)

concentrations at capture. Elevated plasma CK levels are an indication of muscular

damage associated with capture myopathy. My hypothesis was that northem bobwhite

plasma CK measured at capture would be negatively associated with subsequent survival.

Plasma CK of birds differed due to sampling time/location (P = 0.002), but not capture

method (P = 0.870) or year (P = 0.222). Average plasma CK of groups of birds

processed in the lab (9,290 ± 1,000 u/L [x± SE], « = 13) was almost twice that of groups

processed in the field (5,240 ±510 u/L, n = 38). For birds sampled in the field, plasma

CK was negatively associated with probability of survival to 8, 12, 16, and 20 weeks (P <

0.1), but not with probability of survival to 4 weeks (P = 0.773). The probability of a bird

surviving decreased by 14%) with each 1,000 u/L increase of plasma CK. Plasma CK was

not associated v^th probability of survival to any time period for birds sampled after

transport to a field lab (P > 0.5). Survival estimates differed between birds having high

versus low plasma CK (Z= 1.58, P = 0.057; 54.8 ± 8.2% vs. 72.4 ± 7.5%). If plasma CK

accurately indicated severity of capture myopathy, then capture myopathy resulted in

^ 12.2%o negative bias of the survival estimate (63.6 ± 5.7%o) calculated using all of the

data. These results suggest that demographic estimates based on captured animals are

biased. Plasma CK may be a useftal tool to identify individual animals whose probability

of survival has been influenced by capture-related trauma, and should thus be removed

from analyses.

30

Introduction

Much of our knowledge of wildlife has been gained through studies that capture

animals, mark or radiomark them, and then monitor their fate. Biologists have long

realized that the capture event could alter the fate of animals and have attempted to

account for this effect. For example, mark-recapture models have been developed for

situations where the marking event reduces short-term survival of the animal (Pollock et

al. 1990). In radiotelemetry studies, data from animals that die within some specified

interval after capture are often censored (Cox and Afton 1998). For animals that survive

this interval, the assumption is made that subsequent survival of the radiomarked

population is representative of the population of interest (Pollock et al. 1989). However,

current methods of incorporating potential negative capture effects in demographic

analyses are problematic because they typically assume a short-term effect that is

constant for all animals.

A common injury to wdld animals that results from capture is a condition referred to

as capture myopathy. The condition is characterized by acute degeneration of muscle

tissue resulting from intense muscular exertion and trauma caused by a chase, stmggle, or

transport (Hulland 1993). Low pH vsdthin muscle tissue resulting from the buildup of

lactic acid from anaerobic glycolysis results in increased cell membrane permeability and

cell lysis (Wobeser 1997:218). This releases cellular enzymes, such as CK to the blood,

and elevated levels of these enzymes in serum or plasma are indicative of skeletal or

cardiac muscle damage (Bollinger et al. 1989). Scarring of skeletal and cardiac muscles

may occur, which can contribute to death weeks or months after the causative event

(Hulland 1993). The degree of capture myopathy among individuals captured under

similar conditions, as indicated by semm or plasma concentrations of cellular enzymes,

can vary greatly (Bollinger et al. 1989). Thus, assumptions that subsequent survival of

captured animals is representative of the population of interest or that capture effects are

short-term and constant for all capture animals may be violated if capture myopathy

occurs.

31

Studies of capture myopathy in wild birds have involved necropsy and examination

of muscular lesions (Spraker et al. 1987) and measurement of serum enzymes, usually

CK and aspartate aminotransferase (Bollinger et al. 1989, Dabbert and Powell 1993).

Additionally, researchers have associated particular capture and handling techniques with

subsequent survival, and speculated that the differences in survival rates were likely due

to differences in the number of individuals predated as a result of debilitating effects of

capture myopathy (Cox and Afton 1998). However, no studies have measured serum or

plasma enzymes to determine if higher levels were associated with increased mortality.

The objective of this study was to assess whether concentration of plasma CK, an

indicator of capture myopathy, was associated with survival of radiomarked northem

bobwhite.

Methods

Northem bobwhites were captured from January through March of 1997 and 1998

using walk-in funnel traps (Smith et al. 1981) baited with milo and by netting coveys at

night. Traps were checked traps twice per day, about 4 hr after sunrise and 30 min before

sunset, and coveys were netted 2-4 hr after sunset. I could not estimate how long birds

had been trapped, but trap checks were timed to coincide with the end of morning and

evening northem bobwhite activity periods. Birds were immediately removed from traps

or the net and placed them into ventilated sacks or cardboard boxes. Each group of

captured birds was processed either immediately in the field or after transport to a lab

about 11 km away. Birds processed in the field were typically handled <10 min and then

released immediately. In contrast, birds transported to the lab were held for various times

before processing, often held overnight, and then released back at the capture location.

All hens were weighed, marked with an aluminum legband (National Band and Tag

Co., Newport, Kentucky), radiomarked, bled, and released. Each radio transmitter

weighed 6.5 g, had a mortality sensor, and was a necklace style with no body loop

(American Wildlife Enterprises, Montacello, Florida). Each hen was bled by pricking the

ulnar or brachial vein with a 25-gauge needle and collecting blood into heparinized

32

capillary tubes. Capillary tubes were sealed with clay, transported to a field lab, and

centrifuged. Tubes were then scored with a diamond-tipped pen and broken immediately

above the white blood cell layer. Plasma was pipetted into cryovials, frozen initially at

-20 °C for <4 months and then -84 °C for < 18 months, and shipped on dry ice to a

commercial laboratory (Veterinary Associates Laboratory, Edmond, Oklahoma) for

measurement of CK using a Technicon RA-1000® clinical chemistry analyzer (Bayer

Corporation, Pittsburgh, Pennsylvania). Hemolytic and lipemic samples were discarded

due to potential measurement errors (Alleman 1990). Radiomarked birds were monitored

for survival 1-3 times per week from capture until July or August of each year.

Survival rates were calculated using the staggered-entry Kaplan-Meier model

(Pollock et al. 1989) of program BCAPLAN (T. Kulowiec, Missouri Department of

Conservation, Columbia, Missouri). Precision of the survival estimator is poor when

based on <20 individuals (Pollock et al. 1989), so calculations were limited to periods

when ^20 birds were radiomarked. This resulted in survival estimates for 2 165-day

periods: 10 Febmary 1997 through 24 July 1997 and 14 Febmary 1998 through 28 July

1998.

The effects of capture method (trap or net), processing location (field or lab), and

year (1997 or 1998) on levels of plasma CK were assessed using analysis of variance.

The levels of CK from each group of birds captured at a single time were probably

correlated, violating the assumption for analysis of variance of independent error terms

(Sokal and Rohlf 1995:393). Thus, each group of birds captured during a single capture

event was used as the experimental unit and average CK for each group of birds as the

response variable.

The relationship between survival and plasma CK was assessed using logistic

regression. Survival was assessed at 4, 8, 12, 16, 20, and 24 weeks. Individual birds

were used as the experimental units because survival probabilities were probably

uncorrelated with capture events. I tested for a linear relationship between the logit of

survival and plasma CK using a fiill/reduced model likelihood ratio chi-squared test (SAS

33

Institute 1989). Analyses were performed using PROC LOGISTIC of version 6.12 of

SAS (SAS Institute, Gary, North Carolina).

The assumptions of normality and homoscedasticity were assessed where

appropriate using the Shapiro-Wilk (Shapiro and Wilk 1965) and Levene (1960) tests,

respectively. Analyses were conducted using SPSS for Windows, release 6.0 (SPSS,

Chicago, Illinois), unless specified otherwise. Statistical significance was determined

using an a-Ievel of 0.1 rather than the traditional a-Ievel of 0.05 to reduce the

probability of Type II error (Skalski and Robson 1992:21-22). Means are reported ± 1

standard error.

Results

I captured, bled, radiomarked, and monitored 104 northem bobwhite hens (48 in

1997 and 56 in 1998). Survival of hens was 0.525 ± 0.079 from 10 Febmary to 24 July

1997 and 0.625 ± 0.078 from 14 Febmary to 28 July 1998.

Levels of plasma CK with respect to capture method, processing location, and year

were homoscedastic but nonnormal. Because ANOVA is robust to nonnormality (Zar

1996:187), I chose not to transform data or use nonparametric alternatives. Plasma CK

was influenced by processing location (F, 47 = 10.63, P = 0.002), but not capture method

(F, 47 = 0.03, P = 0.870) or year (F, 47 = 1.53, P = 0.222). Average plasma CK of groups

of birds processed in the lab (9,290 ± 1,000 u/L, n=\3) was almost twice that of groups

processed in the field (5,240 ±510 u/L, n = 38). Because of the observed difference in

plasma CK of birds processed in the field versus the lab, I assessed the relationship

between plasma CK and survival separately for each group. For birds that were

transported back to the lab for processing, plasma CK was not related to the logit of

survival for any interval (Table 4.1). For birds processed in the field, plasma CK was

related to the logit of survival measured to 8, 12, 16, and 20 weeks (Table 4.1). The

probability of a bird surviving to 16 weeks decreased by 14% with each 1,000 u/L

increase of plasma CK (Figure 4.1).

34

Table 4.1. Tests of linear relationships between the logit of survival to 6 time periods and plasma CK sampled either after transport to a laboratory or immediately in the field.

Sampling location Survival interval (weeks)

Laboratory

Field

4

8

12

16

20

24

4

8

12

16

20

24

n /

32

31

30

30

28

16

72

72

70

67

59

43

<0.01

0.25

0.40

0.43

0.26

0.03

0.08

3.00

2.97

3.59

3.42

0.04

0.957

0.616

0.528

0.514

0.613

0.874

0.773

0.083

0.085

0.058

0.065

0.845

35

> >

CO

1.0

0.8

0.6

0.4

0.2

0.0

• • • • • • • •

0 5,000 10,000

Plasma CK (u/L)

15,000

Figure 4.1. Relationship between survival to 16 weeks of northem bobwhite hens (n 61) and concentration of plasma CK for birds sampled in the field. The logit of survival decreased with increasing levels of plasma CK (P = 0.058). The equation depicted by the line is y = e ' 23-000013ix)/( i + d 23-0.00013 Ix)

) .

36

Discussion

Plasma CK was negatively related to survival of northem bobwhites when plasma

CK was sampled immediately after birds were removed from the traps or net. These

results suggest that the effects of capture do not impact all animals equally, an

assumption of models that attempt to incorporate capture effects. Plasma CK may be a

useful diagnostic tool for determining the degree that each animal was impacted by

capture, and the subsequent effect on its survival probability. For birds handled in the

field, I was surprised to find a relationship between plasma CK and survival measured to

every interval from 8 to 20 weeks, but not to 4 weeks (Table 4.1). This suggests that the

effect of capture myopathy on northem bobwhite survival may take >4 weeks to manifest

itself. A conditioning period >4 weeks would be required to account for these effects

using current methods for studies of radiomarked animals; few researchers will find this

feasible given the cost of field studies and the current battery life of transmitters for

avians.

To explore the effect of plasma CK on survival estimates, I tested the hypothesis that

survival of northem bobwhites with high plasma CK was lower than survival of northem

bobwhites with low plasma CK. Hill and Murray (1987) estimated the normal bounds of

plasma CK to be 90-1,590 u/L. Only 1 \% of the birds sampled in the field fell within

this range. Median plasma CK of birds sampled in the field during both years was

3,760 u/L. I used this value to divide these birds into low and high plasma CK groups. I

combined both years of data and restricted the comparison to the interval when ^20

radiomarked birds were monitored (Pollock et al. 1989); this was the 118-day period from

16 Febmary to 14 June. Survival was significantiy lower for the high CK group (1-tailed

Z-test, Pollock et al. 1989; Z= 1.58, P = 0.057; 54.8 ± 8.2% vs. 72.4 ± 7.5%). Estimated

survival during this interval using the combined data was 63.6 ± 5.7%). Assuming plasma

CK accurately indicated severity of capture myopathy and that birds in the low CK group

were unaffected by capture myopathy, then capture myopathy resulted in a -12.2% bias of

the survival estimate calculated using all of the data. Because only 8 of 36 birds in the

37

low CK group had plasma CK levels wdthin the normal range (Hill and Murray 1987), it

is doubtful that these birds were completely unaffected by capture myopathy. Thus, this

estimate of negative bias due to capture myopathy may be conservative. Also, other

factors may have caused further bias in the survival estimate. For example, Osborne et al.

(1997) found that simulated radio transmitters caused birds to lose body mass and lipids,

and thus could potentially lower survival. These other potential effects could not be

assessed in this study. Researchers must recognize that the assumption that capture and

radiomarking animals has either no effect or a constant, short-term effect may be severely

violated.

The failure to find a relationship between survival and plasma CK for birds

sampled in the lab may have been due to the smaller sample size or the lack of

consistency in sampling time for this group. The rise in plasma CK caused by physical

exercise or capture of animals is time dependent. Tripp and Schmitz (1982) found that

CK levels in domestic turkeys were unchanged immediately following 15 min of

continuous exercise, but rose about 5-fold during the next 4 hr and remained at that level

for at least another 24 hr. Dabbert and Powell (1993) found that CK levels in domestic

mallards (Anas platyrhynchos) sampled 45 min following the initiation of capture and

transport was 6-fold that of baseline levels. Thus, plasma CK probably begins to rise

yvithin 15 to 45 min of capture. The difference in plasma CK between birds handled in

the field and lab was probably largely due to the greater time that elapsed between

capture and sampling. The difference may also have partially been due to the effect of

transportation; the tme cause cannot be determined because sampling time and

transportation are confounded. This example illustrates that handling and sampling times

must be consistent for semm or plasma enzyme results to be of any value.

38

Literature Cited

Alleman, A. R. 1990. The effects of hemolysis and lipemia on serum biochemical constituents. Veterinary Medicine 85:1272-1284.

Bollinger, T., G. Wobeser, R. G. Clark, D. J. Nieman, and J. R. Smith. 1989. Concentration of creatine kinase and aspartate aminotransferase in the blood of wild mallards following capture by three methods for banding. Joumal of Wildlife Diseases 25:225-231.

Cox, R. R., Jr., and A. D. Afton. 1998. Effects of capture and handling on survival of female northem pintails. Joumal of Field Ornithology 69:276-287.

Dabbert, C. B., and K. C. Powell. 1993. Semm enzymes as indicators of capture myopathy in mallards (Anas platyrhynchos). Joumal of Wildlife Diseases 29:304-309.

Hill, E. F,, and H. C. Murray. 1987. Seasonal variation in diagnostic enzymes and biochemical constituents of captive northem bobwhites and passerines. Comparative Biochemistry and Physiology 87B:933-940.

Hulland, T.J. 1993. Muscle and tendon. Pages 183-265 m K. V. F. Jubb, P. C. Kennedy, and N. Palmer, editors. Pathology of domestic animals. Fourth edition. Volume 1. Academic Press, New York, New York.

Levene, H. 1960. Robust tests for equality of variances. Pages 278-292 in I. Olkin, editor. Contributions to probability and statistics. Stanford University Press, Palo Alto, California.

Osbome, D. A., B. J. Frawley, and H. P. Weeks, Jr. 1997. Effects of radio tags on captive northem bobwhite (Colinus virginianus) body composition and survival. American Midland Naturalist 137:213-224.

Pollock, K. H., S. R. Winterstein, C. M. Bunck, and P. D. Curtis. 1989. Survival analysis in telemetry studies: the staggered entry design. Joumal of Wildlife Management 53:7-15.

, J. D. Nichols, C. Brownie, and J. E. Hines. 1990. Statistical inference for capture-recapture experiments. Wildlife Monograph 107:1-97.

SAS Institute. 1989. SAS/STAT user's guide. Version 6. Fourth edition. Volume 2. SAS Institute, Gary, North Carolina.

Shapiro, S. S., and M. B. Wilk. 1965. An analysis of variance test for normality (complete samples). Biometrika 52:591-611.

Skalski, J. R., and D. S. Robson. 1992. Techniques for wildlife investigations: design and analysis of capmre data. Academic Press, San Diego, California.

39

Smith, H. D., F. A. Stormer, and R. D. Godfrey, Jr. 1981. A collapsible quail trap. U.S. Forest Service Research Note RM-400.

Sokal, R. R., and F. J. Rohlf 1995. Biometry. Third edition. W. H. Freeman and Company, New York, New York.

Spraker, T. R., W. J. Adrian, and W. R. Lance. 1987. Capture myopathy in wild mrkeys (Meleagris gallopavo) following trapping, handling, and transportation in Colorado. Joumal of Wildlife Diseases 23:447-453.

Tripp, M. J., and J. A. Schmitz. 1982. Influence of physical exercise on plasma creatine kinase activity in healthy and dystrophic turkeys and sheep. American Joumal of Veterinary Research 43:2220-2223.

Wobeser, G. A. 1997. Diseasesof wild waterfowl. Second edition. Plenum Press, New York, New York.

Zar, J. H. 1996. Biostatistical analysis. Third edition. Prentice Hall, Upper Saddle River, New Jersey.

40

CHAPTER V

RELATIONSHIP BETWEEN PLASMA TRIGLYCERIDE, BODY MASS,

AND REPRODUCTION OF NORTHERN BOBWHITES

Abstract

The earliest potential initiation of northem bobwhite (Colinus virginianus)

reproduction is limited by photoperiod. Secondary factors such as lipid reserves, diet, and

stress often limit the beginning of northem bobwhite reproduction, potentially reducing

reproductive success and causing a shorter reproductive season. I measured late winter

body masses and plasma triglycerides of wild northem bobwhites and subsequent

reproductive timing and effort in 1997 and 1998 on the Coastal Prairie of Texas. Using

body mass and plasma triglyceride level as indices of body fat, I tested the hypothesis that

the onset of reproduction and first clutch size was influenced by late winter lipid reserves.

Northem bobwhite plasma triglycerides were higher (P < 0.001) and more variable (P =

0.019) in 1998, and nesting began 15 ± 1.6 days (x± SE) earlier than in 1997. However,

within each year no combination of body mass and triglycerides was associated with

timing of nesting or size of first clutch (P > 0.1). In addition, body masses were not

correlated with plasma triglycerides (P > 0.1), My findings suggest that individual

plasma triglyceride level and body mass are unsuitable variables for assessing within-

population differences in reproductive timing. However, mean plasma triglycerides for a

population may be useful for assessing differences in reproductive timing among years

and locations. The relationship between triglycerides and hormones directly affecting

gonadal recmdescence, such as luteinizing hormone (LH) and prolactin, is uncertain for

wild northem bobwhites. Thus, future studies should assess causes and patterns of

change in these hormones.

Introduction

The initiation of reproduction in birds is controlled by the endocrine system (Bahr

and Johnson 1991). Photoperiod has an overriding influence on reproduction in many

41

species, including northem bobwhite (Kirkpatrick and Leopold 1952), by regulating

hypothalamic secretions of follicle stimulating hormone (FSH) and LH (Famer and

Follett 1979, Carey 1996). After minimum photoperiod is available to birds, secondary

factors may influence hormone levels and determine when individuals begin to reproduce.

Physiological condition, diet, and stress are 3 interrelated secondary factors influencing

northem bobwhite reproduction. For example, water deprivation in northem bobwhites

decreases semm progesterone, delays or prevents egg production, decreases clutch size,

and causes smaller eggs (Cain and Lien 1985, Koerth and Guthery 1991, Giuliano et al.

1995). Insufficient dietary energy increases semm corticosterone and decreases northem

bobwhite egg production (Giuliano et al. 1996). And, lower body mass in late winter may

delay the onset of egg laying in pheasants, probably due to levels of stress and

reproductive hormones (Gates and Woehler 1968). Physiological condition, diet, and

stress are probably strongly correlated among wild birds at any particular time and place

because these factors are largely dependent on environmental conditions. However,

differences in condition, diet, and stress within a population may explain within-year

variability in the timing of reproduction and reproductive effort.

Late winter lipid reserves probably influence future reproduction (Gates and

Woehler 1968), although lipid reserves are less important to reproduction of grouses and

presumably other galliforms as compared to waterfowl (Thomas 1988). Lipid reserves

could increase size of the first clutch by allowing a female to maintain and recmit more

follicles for ovulation, largely through the action of FSH (Carey 1996). Lipid reserves are

correlated with northem bobwhite body masses during winter (Rebel 1972, Frawley et al.

1999), and with both body masses and plasma triglycerides when measured in other

avians (Bacon et al. 1989, Dabbert et al. 1997). Thus, late winter body masses and

plasma triglycerides may explain within-year variability of reproductive timing and first

clutch sizes. I tested the hypothesis that the onset of reproduction and first clutch size of

northem bobwhites was influenced by late winter lipid reserves, using body masses and

plasma triglyceride levels as indices of body fat.

42

Methods

Methods for capturing and monitoring northem bobwhites were described in Chapter

II. Briefly, female northem bobwhites were captured from 1 Febmary through 23 April in

1997 and 1998 primarily with walk-in funnel traps (Smith et al. 1981) baited with milo.

All hens were weighed, marked with an aluminum legband (National Band and Tag Co.,

Newport, Kentucky), radiomarked (American Wildlife Enterprises, Montacello, Florida),

bled, and released. Each hen was bled by pricking the ulnar or brachial vein with a 25-

gauge needle and collecting blood into heparinized capillary tubes. Capillary tubes were

sealed with clay, transported to a field lab, and centrifuged. Tubes were then scored with

a diamond-tipped pen and broken immediately above the white blood cell layer. Plasma

was pipetted into cryovials, frozen initially at -20°C and then -84°C, and shipped on dry

ice to a commercial laboratory (Veterinary Associates Laboratory, Edmond, Oklahoma)

for measurement of triglycerides using a Technicon RA-1000® clinical chemistry

analyzer (Bayer Corporation, Pittsburgh, Pennsylvania). Hemolytic and lipemic samples

were discarded due to potential measurement errors (Alleman 1990).

From late April until mid-July, most radiomarked hens were located at least once

every other day to find nests. Each bird was approached on foot until it flushed or moved

away, was observed, or had been circled by the researcher, indicating that it might be on a

nest. Two sets of stake flags were placed at a distance of 5 and 10 m from the nest

oriented so that each set was aligned with the nest. The position of the nest could be

determined on subsequent visits by sighting down both sets of stake flags and

determining the point where the 2 lines intersected. This marking method was used so

that predators investigating the stake flags were not drawn directly to the nest. Hens were

never intentionally flushed from nests and rarely flushed inadvertently. To confirm the

nest location and clutch size, the nest site was checked at times when the hen might be

away from the nest feeding. When the nest was unoccupied, the number of eggs was

recorded. The date that eggs in a nest began to be incubated was estimated by averaging

the last date that a hen was not found on a nest and the first date that a hen was found

43

incubating eggs on a nest. The date that nesting began was estimated by subtracting the

number of eggs in a nest from the estimated date that incubation of eggs began.

Data Analysis

The value of late-winter body mass and plasma triglycerides as predictors of first

clutch size and date of first nest initiation was assessed using multiple linear regression.

Body masses and plasma triglycerides were recorded from birds captured 6-16 Febmary

1997 and 4 Febmary-7 March 1998. Records for 5 nests initiated >30 days after the

earliest nests each year were censored because it was suspected that the initial nests for

these birds were not detected. First nests would probably not have been detected if they

were depredated before incubation began or if the male incubated the eggs.

Residuals were tested for normality using the Shapiro-Wilk test (Shapiro and Wilk

1965). Homoscedasticity and linearity were assessed by examining plots of residual and

predicted values. Analyses were conducted using SPSS for Windows, release 6.0

(Nomsis 1993). Statistical significance was determined with a = 0.05. Means are

reported ± 1 SE.

Results

Body masses, plasma triglyceride levels, dates of nest initiation, and sizes of first

clutches were obtained for 17 northem bobwhite hens in 1997 and 23 in 1998 (Table 5.1).

Northem bobwhites began nesting 15 ± 1.6 days earlier in 1998 ( 35 = 6.70, P < 0.001).

Clutch sizes did not differ between years (t-^^ = 0.86, P = 0.394). The mean and variance

of plasma triglycerides were higher in 1998 (mean: t-test for unequal variances, 30 =

-5.53, P < 0.001; variance: Levene's [1960] test, F, 33 = 5.99, P = 0.019). Body masses

were lower in 1998 (t-test for equal variances, 35 = 2.09, P = 0.043). Plasma triglycerides

were not correlated with body mass either year (1997: R^ = 0.01, F, ,5 = 0.20, P = 0.660;

1998: R^ = 0.12, F, 2, = 2.87, P = 0.105; Figure 5.1).

44

Table 5.1. First clutch sizes, dates of nest initiation, late winter body masses (g), and plasma triglyceride levels (mg/dL) of northem bobwhites in 1997 (n= \1) and 1998 (n = 23), Refugio County, Texas.

First clutch size

Date of nest initiation

Body mass

Plasma triglyceride

1997

X

15.7

3 May

167.7

122.2

SE

0.3

1.9

2.6

9.6

1998

X

15.2

18 April

161.1

250.1

SE

0.4

1.4

1.9

21.0

Tests for relationships among variables were conducted separately for each year

because birds began reproducing earlier in 1998. No combination of body mass and

triglycerides successfully predicted date of nest initiation or first clutch size (Table 5.2).

Assumptions of normality and homoscedasticity were violated for analyses of clutch size

in 1998 due to a clutch of 22 eggs. The analysis was repeated with this record removed

with identical results.

Discussion

Late winter body mass and plasma triglycerides were not related to date of nest

initiation or first clutch size either year. This contradicts patterns found in turkeys

(Meleagris gallopavo; Badyaev et al, 1996), where body mass in Febmary-March and

nest initiation date were the best predictors of clutch size of first nests. However, in the

turkey study (Badyaev et al. 1996), the partial r for female body mass was 0,10. Thus,

turkey body mass accounted for only an additional !%> of the variation in clutch sizes

given the other variable in their model. If body mass influenced clutch size this weakly in

my study, I could not have detected it. Both studies support earlier conclusions that

galliform reproduction relies very little on lipid reserves (Thomas 1988),

45

^ 500 _ i 5 P 400 • ^^

(0 ^ 300 '&—

0 ^ 200 D)

^ 100

n u

•

• •

^

•

•

•

•

•

1997

•

• • •

1

3- 500

I' 400 (/)

^ 300

^ 200

• 100 0

-

-

•

•

-

• •

1

•

•

• •

• •

•

• ••

•

•

•

• •

• •

• •

1

1998

•

1

140 150 160 170 180 190

Body Mass (g)

Figure 5,1. Lack of linear relation between plasma triglycerides and body mass of northem bobwhites in 1997 and 1998, Reftigio County, Texas (1997: R^ 0.01, F , ,5 = 0.20, P = 0.660; 1998: R^ = 0.12, F , 2, = 2.87, P = 0.105).

46

Table 5.2. R^ and significance of late winter body mass (BM) and plasma triglycerides (TRIG) as predictors of first clutch size and date of nest initiation for northem bobwhites in Refugio County, Texas, 1997-1998.

Dependent variable Year Independent variables R'

First clutch size

Date of nest initiation

1997 BM + TRIG

BM

TRIG

1998 BM + TRIG

BM

TRIG

1997 BM + TRIG

BM

TRIG

1998 BM + TRIG

BM

TRIG

0.19

0.12

0.09

0.03

0.00

0.03

0.03

0.02

0.02

0.08

0.03

0.03

0.235

0.172

0.249

0.752

0.970

0.466

0.794

0.582

0.634

0.414

0.463

0.435

In 1997, plasma triglyceride levels from my study were similar to winter plasma

triglyceride levels reported for northem bobwhites housed in outdoor pens and fed a

commercial gamebird feed ad libitum (Hill and Murray 1987). However, triglyceride

levels in 1998 were twice as high and twice as variable. I examined the relationship

between sampling date and plasma triglycerides in 1998 to determine if my sampling

interval extended into the beginning of the breeding season when triglycerides increase 3-

to 6-fold (Hill and Murray 1987). If this occurred, I predicted that plasma triglycerides

would increase during the 31-day sampling interval in 1998. However, I found no linear

relationship between plasma triglycerides and sampling date (F, 21 = 0.5, P = 0.507;

Figure 5.2). I suspected that differences in sampling locations also may have caused the

observed annual differences. In 1997, 13 of 17 samples were obtained from birds

captured within a 1-km radius. In 1998, samples were obtained from 8 locations

47

separated by >2 km, and the largest number of samples from any single location was 7.

To test whether the differences in the means and variability of plasma triglycerides were

due to location, I compared the samples collected in 1997 from a single location with 4

samples collected from that location in 1998, Surprisingly, the means and trends for 1997

and 1998 of plasma triglyceride levels and body masses from this single location were

nearly identical to that of all locations (Table 5.3). Thus, I do not know why triglyceride

levels were so much greater and variable in 1998.

Table 5.3. Comparison of northem bobwhite late winter body masses (g) and plasma triglyceride levels (mg/dL) in 1997 and 1998 between a single pasture (Willow Lake Trap [WLT], « = 13 and 4, respectively) and all locations (n = 17 and 23, respectively), Refugio County, Texas. Means are followed by standard deviations (SD).

1997 1998

WLT All WLT All

SD X SD X SD X SD

Body mass 166.8 11.2

Plasma triglyceride 125.8 38.8

167.7

122.2

10.8

39.5

156.3

262.0

8.9

190.5

161.1

250.1

9.3

100.9

Higher plasma triglycerides in 1998 were associated with earlier reproduction.

However, I found no relationship between individual plasma triglyceride levels and date

of reproduction. I suspect that northem bobwhites in my study were differentially

affected by numerous extrinsic factors, such as differences in diet caused by habitat

quality and disturbance by predators. In addition, some extrinsic factors probably

affected all of the birds to nearly the same extent within a given year; these effects would

include winter severity and precipitation. My resuUs suggest that plasma triglycerides

may be useful as an indicator of population trends in timing of reproduction, but not for

predicting reproductive timing of individuals.

48

600

2/1/98 2/8/98 2/15/98 2/22/98 3/1/98 3/8/98 Sampling Date

Figure 5.2 Lack of linear relationship between plasma triglycerides of northem bobwhites and sampling date in 1998, Refugio County, Texas (F, 21 = 0.5, P = 0.507).

Body mass and plasma triglycerides were not positively related (Figure 5.1). In fact,

the pattern in 1998, though not statistically significant, was a negative relationship.

These results indicate a need for a better understanding of how body mass, plasma

triglycerides, and body fat vary among wild birds subjected to stressftjl late winter

conditions. In addition, better information is needed on the timing of changes in these

variables as photoperiod increases to levels that can stimulate gonadal recmdescence.

Our ability to predict northem bobwhite production is hampered by an incomplete

understanding of the factors controlling reproductive timing and effort. Earlier nesting of

galliforms is advantageous due to lower nest predation, higher hatchability, and greater

potential for renesting (Klimstra and Roseberry 1975, Guthery et al. 1988, Badyaev et al.

49

1996, but see Lehmann 1946), Photoperiod limits the potential reproductive period in

northem bobwhites (Kirkpatrick and Leopold 1952), but secondary factors such as lipid

reserves, diet, and stress can constrain the realized reproductive period (Gates and

Woehler 1968, Cain and Lien 1985, Koerth and Guthery 1991, Giuliano et al, 1995,

1996). My findings indicate that plasma triglycerides and body masses may be useful for

comparing populations among years and locations, but these variables do not explain

within-population variation in reproductive timing and effort. Future research on

reproduction of northem bobwhites should measure hormones more closely linked to

gonadal recrudescence and ovulation and inhibitory effects of stress hormones such as

corticosterone. Research on reproduction of wild northem bobwhite can build on the

excellent foundation of recent laboratory studies that measured reproductive and stress

hormones (Cain and Lien 1985, Giuliano et al. 1995, 1996).

50

Literature Cited

Alleman, A, R. 1990. The effects of hemolysis and lipemia on semm biochemical constituents. Veterinary Medicine 85:1272-1284.

Bacon, W. L,, K, E, Nestor, and E, C, Naber. 1989. Prediction of carcass composition of turkeys by blood lipids. Poultry Science 681282-1288.

Badyaev, A. V,, T. E. Martin, and W. J. Etges. 1996, Habitat sampling and habitat selection by female wild turkeys: ecological correlates and reproductive consequences. Auk 113:636-646.

Bahr, J. M., and P. A. Johnson. 1991. Reproduction in poultry. Pages 555-575 in P. T. Cupps, editor. Reproduction in domestic animals. Fourth edition. Harcourt Brace Jovanovich, San Diego, California.

Cain, J. R., and R. J. Lien. 1985. A model for drought inhibition of bobwhite quail (Colinus virginianus) reproductive systems. Comparative Biochemistry and Physiology 82A:925-930.

Carey, C, 1996. Female reproductive energetics. Pages 324-374 m C. Carey, editor. Avian energetics and nutritional ecology. Chapman & Hall, New York, New York.

Dabbert, C, B., T, E. Martin, and K. C. Powell. 1997. Use of body measurements and semm metabolites to estimate the nutritional status of mallards wintering in the Mississippi Alluvial Valley, USA. Joumal of Wildlife Diseases 33:57-63.

Famer, D. S., and B. K. Follett. 1979. Reproductive periodicity in birds. Pages 829-872 in E. J. W. Barrington, editor. Hormones and evolution. Volume 2. Academic Press, New York, New York.

Frawley, B. J., D. A. Osbome, H. P. Weeks, Jr., L. W. Burger, Jr., and T. V, Dailey. 1999. Use of total body electrical conductivity to predict northem bobwhite lipid mass. Joumal of Wildlife Management 63:695-704.

Gates, J. M., and E. E. Woehler, 1968. Winter weight loss related to subsequent weights and reproduction in penned pheasant hens. Joumal of Wildlife Management 32:234-247.

Giuliano, W, M,, R. S. Lutz, and R Patino. 1995. A simulated insect diet as a water source for quail: effects on body mass and reproduction. Comparative Biochemistry and Physiology 111 A:299-302.

, , and . 1996, Reproductive responses of adult female northem bobwhite and scaled quail to nutritional stress. Joumal of Wildlife Management 60:302-309.

Guthery, F. S., N. E. Koerth, and D. S. Smith. 1988. Reproduction of northem bobwhites in semiarid environments. Joumal of Wildlife Management 52:144-149,

51

Hill, E. C , and H. C. Murray. 1987. Seasonal variation in diagnostic enzymes and biochemical constituents of captive northem bobwhites and passerines. Comparative Biochemistry and Physiology 87B:933-940,

Kirkpatrick, C. M., and A. C. Leopold. 1952. The role of darkness in sexual activity of thequaiL Science 116:280-281.

Klimstra, W. D., and J. L. Roseberry. 1975. Nesting ecology of the bobwhite in southern Illinois. Wildlife Monographs 41:1-37.

Koerth, N. E,, and F. S. Guthery. 1991. Water restriction effects on northem bobwhite reproduction. Joumal of Wildlife Management 55:132-137.

Lehmann, V. W. 1946. Bobwhite quail reproduction on southwestem Texas. Joumal of Wildlife Management 10:111-123.

Levene, H. 1960. Robust tests for equality of variances. Pages 278-292 in I. Olkin, editor. Contributions to probability and statistics. Stanford University Press, Palo Alto, California.

Nomsis, M. J. 1993. SPSS® for Windows: base system user's guide. Release 6.0. SPSS, Chicago, Illinois.

Robel, R. J. 1972. Body fat content of bobwhites in relation to food plantings in Kansas. National Bobwhite Quail Symposium 1:139-149.

Shapiro, S. S., and M. B. Wilk. 1965. An analysis of variance test for normality (complete samples). Biometrika 52:591-611.

Smith, H. D., F. A. Stormer, and R. D. Godfrey, Jr. 1981. A collapsible quail trap. U.S. Forest Service Research Note RM-400.

Thomas, V. G. 1988. Body conditions, ovarian hierarchies, and their relation to egg formation in Anseriform and Galliform species. Pages 353-363 in H. Oullet, editor. Acta XIX Congressus Intemationalis Omithologici. University of Ottawa Press, Ottawa, Canada.

52